[🐛 Bug]: [py] BiDi Network handlers fail on data URLs

[cgoldberg]

Description

When I add a request handler via BiDi, I am unable to .continue_request() when the request contains a data URL. (I only tried this with Chrome, so I'm not sure how other browsers handle this).

"Data URLs" contain embedded data rather than a normal network location. They are common for embedding fonts, images, etc.

https://developer.mozilla.org/en-US/docs/Web/URI/Reference/Schemes/data

To reproduce: add a request handler and then navigate to a page that will make subsequent requests for a data URL.

Reproducible Code

from selenium import webdriver

def log_urls(request):
    print(f"Intercepted URL: {request.url}")
    request.continue_request()

options = webdriver.ChromeOptions()
options.enable_bidi = True
driver = webdriver.Chrome(options=options)
driver.network.add_request_handler("before_request", log_urls)
driver.browsing_context.navigate(
    url="https://selenium.dev", context=driver.current_window_handle
)

Debugging Logs

File "/home/cgoldberg/code/venv/lib/python3.13/site-packages/selenium/webdriver/common/bidi/network.py", line 143, in _callback
    callback(request)
    ~~~~~~~~^^^^^^^^^
  File "/home/cgoldberg/.pyenv/versions/3.13.7/lib/python3.13/threading.py", line 1043, in _bootstrap_inner
    self.run()
    ~~~~~~~~^^
  File "/home/cgoldberg/code/./foo.py", line 11, in log_urls
    request.continue_request()
    ~~~~~~~~~~~~~~~~~~~~~~~~^^
  File "/home/cgoldberg/code/venv/lib/python3.13/site-packages/selenium/webdriver/common/bidi/network.py", line 315, in continue_request
    self.network.conn.execute(command_builder("network.continueRequest", params))
    ~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/cgoldberg/.pyenv/versions/3.13.7/lib/python3.13/threading.py", line 994, in run
    self._target(*self._args, **self._kwargs)
    ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/cgoldberg/code/venv/lib/python3.13/site-packages/selenium/webdriver/remote/websocket_connection.py", line 73, in execute
    raise WebDriverException(error_msg)
  File "/home/cgoldberg/code/venv/lib/python3.13/site-packages/selenium/webdriver/remote/websocket_connection.py", line 86, in _callback
    callback(event.from_json(params))
    ~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^
selenium.common.exceptions.WebDriverException: Message: no such request: Network request with ID '10788.49' doesn't exist

[selenium-ci]

@cgoldberg, thank you for creating this issue. We will troubleshoot it as soon as we can.

Selenium Triage Team: remember to follow the Triage Guide

[cgoldberg]

@navin772 any thoughts on this?

[navin772]

This gives a similar error in firefox too - no such request: Blocked request with id X not found. My best guess is that network module isn't recognizing these data: URLs as network requests since they are just inline data.

This is what I think is going on:

1. BiDi intercepts the data: resource, this happens when we add the before_request handler.
2. The browser processes the data and this does not requires/utilizes anything network related since all the data is inline.
3. Then we try to do continue_request()but the data: one was already processed so we get the error - request with ID X not found.

I don't know if its a BiDi/browser side issue. I think the spec hasn't covered this edge case, it intercepts the data: URLs but doesn't takes into account that it will be immediately processed by the browser.

I think this issue is exactly the same - #15553 (comment), the user confirmed the behaviour with firefox, and at that time chrome didn't support continue_request() so it was throwing a diff error.

@whimboo could you confirm if this is the correct behaviour or we need some changes in the bidi spec to cover this edge case?

[whimboo]

Network interception for data: URLs isn't supported yet. See w3c/webdriver-bidi#727 as reference.

[cgoldberg]

I added an xfail test for this in Python: #16281